The trajectory suggests genuine improvement rather than financial engineering, since leverage fell over the same period (D/E 0.53 to 0.13), meaning the ROIC gain is being produced by margin expansion (operating margin 5.6% to 7.9%) rather than debt-funded asset growth. However, the sequential ROIC drop from 13.1% in 2026Q2 to 11.8% in 2026Q3 alongside a fall in asset turnover from 0.52 to 0.47 hints that newly commissioned AI capacity is not yet fully productive, so investors should watch whether the next two quarters restore the upward slope or confirm a plateau.

Roughly 21 of the 44-day improvement comes from DPO expansion, which may indicate Celestica is exercising greater supplier terms power as its order volume grows, but it also raises the question of whether this is a durable structural shift or a timing effect. Receivables efficiency has improved more steadily (DSO 71 to 56 days), which appears consistent with faster hyperscaler billing cycles, while DIO only modestly improved (81 to 70 days), suggesting inventory remains the swing factor investors should monitor if AI component supply tightens again.

Against the 6.48x debt/EBITDA and 9.3x coverage readings in 2025Q1, debt service appears far more comfortable today, and the near-flat absolute debt level ($772M–$1.1B range) implies deleveraging is being driven by EBITDA growth rather than repayment. Given that capital expenditure has risen to 4.7% of revenue, the current 1.78x multiple suggests meaningful headroom to lever if management chooses to fund capacity with debt, so a shift toward higher borrowings would not by itself signal distress.

The gap between the current and quick ratios has narrowed to 0.58x from roughly 0.61x in early periods, which implies inventory still represents a meaningful share of current assets and would be the first balance sheet item to impair under a demand shock. Under stress, the 48-day cash cycle and 1.78x debt/EBITDA would presumably allow the company to fund payables compression without new financing, though the reported cash build in 2026Q3 warrants confirmation before treating the buffer as durable.

The 12-point spread between trailing and forward P/E suggests consensus expects EPS to compound sharply, yet the 60.8x P/FCF versus 33.8x P/E indicates the market is simultaneously discounting weak free cash flow conversion, consistent with capex running at nearly five times depreciation. Relative to EMS peers at 14.57x–20.76x EV/EBITDA, Celestica's 22.62x trailing multiple appears to already embed the AI-infrastructure premium; on a forward 15.03x basis it looks closer to peer norms, so the entry point depends heavily on whether forward EBITDA estimates hold.

The 0.13x debt-to-equity ratio compares favorably with Jabil's 2.22x and Flex's 0.84x, suggesting Celestica's return profile is achieved with materially less balance sheet risk, which appears structurally favorable rather than cyclical given the company's stated preference for organic capacity funding. The key gap versus Jabil's 30.9% ROIC looks structural — Jabil carries higher leverage and a different mix — so investors should not read the spread as underperformance without adjusting for Celestica's heavier current reinvestment phase.